Changing Parameters for Certification Tests:

MediaMatrix allows instructors a relatively broad range of settings for customizing the Chapter Certification Tests that you must require students to take if you adopt our services. It is only through this requirement for testing and your use of these scores in your final grade that you guarantee that students purchase and read the text. But you may choose to handle testing in a variety of ways. Among the most commonly used possibilities are:
1. Use testing which is unsupervised, and thus uses "timed" questions to hedge against cheating by not allowing students time to look up answers in printed copies of the text. A more positive view of timed questions is an emphasis on student "fluency" and the fact that you want that to count as well as "accuracy."

2. Use testing which is supervised. In this case, if you want to stress fluency you may still wish to use "timed" question presentations. If you do not want to stress fluency, but rather would prefer to simulate normal "in-class testing using printed tests," you may wish to merely establish a total time period for testing and allow "return access" or "browsing" of all questions formats.

To establish either of these "timed questions" or "untimed questions" formats, you begin by navigating to the "Certification" screen for work. This is accomplished from any screen by selecting this option from the "Group Parameters" navigation menu at the top of your screen as illustrated below:  

click on picture to go to a full-screen version  

This brings you to the screen where you can select the "style" of your test using the pop-up list selections as illustrated below:

 

click on picture to go to a full-screen version  

From this screen, you may not only set the "style" of your test, but also the specific parameters. Such parameters include:

1. How long students have for each question, if you are using timed questions, or how long they have to take their entire test, if you are using total-test time outs.

2. How many, and what type, questions to use to configure the test. This option allows you to determine the "length" of the test in terms of question items.

Mechanics for these are as simple as selecting a field and retyping the parameters to your desired values. And as the graphic below illustrates, you may set these parameters for the entire class/section (using the "Save Composition for Group" button to register your settings):

 

click on picture to go to a full-screen version  

Alternatively, you may establish different settings for an individual student (who may, for example, be learning disabled and require extended time parameters different from the class). To accomplish this, use the procedure for setting your Control Application to work with a specific student database record ID (as described in the "Setting Individualized Deadlines for a Single Student:" section of the "Certification Deadlines" page of this User's Guide), then register your test parameters only for that student's ID (shown in the lower-right "Student's ID" field) by clicking on the button titled "Send composition to:" as illustrated below:

 

click on picture to go to a full-screen version  

IMPORTANT NOTE ABOUT MEDIAMATRIX ADAPTIVE TESTING: All tiutoring and testing in MediaMatrix is "adaptive." Our definition of this term is a bit different from the emerging literature on this topic. In our system, "adaptive" means that any previous interactions an individual student has with tutoring, quizes, self-assessment, or prior certification attempts will affect the selection of his/her tutoring and test items. We use algorithms which mix 1) unsampled test items with 2) items previously missed (and thus in need of correction), as well as 3) a "base" of previously correct items to guarantee a certain degree of success (if they retain previously learned material, of course). If a student has never taken a test on any topic within a given chapter, then all items are "unsampled" and thus the test is composed of random selections among the chapter. Once a test has been attempted, the student has established a history (or the system has a sample-based "mirror image" of his/her knowledge) and the next test begins with the "mixture" algorithms. We never use the same test twice unless item selection duplication happens by chance, which is extremely unlikely.

Next Topic.....(Tutoring Parameters)